Omnibus: Redis server is stuck
Hey there
We have a GitLab 8.5.5 Omnibus installation and we are having consistent problems with the redis
server:
Errno::EAGAIN (Resource temporarily unavailable - connect(2) for /var/opt/gitlab/redis/redis.socket
It can't access redis
even if the service is on. redis-cli
can't access either.
From our findings, it happens when there is no pipe on the stout
. So, we fix it by listening to its STDOUT.
nohup cat /proc/$(ps aux | grep redis-server | grep -v grep | awk '{ print $2 }')/fd/1 > /dev/null &
Is there a way to configure it to throw its logs to /dev/null
without this piece of bash? logging to a file is also good. gitlab.rb
doesn't seem to make it happen in our case...
Thank you, and I hope this little bash one liner could help any of you guys.